A male infant, weighing 2.75 kg, was born to a 22 year old insulin-dependent diabetic woman, at 36 weeks of gestation. The mother was a carrier of metachromatic leucodystrophy and had a poor obstetric history of four miscarriages and one therapeutic abortion. The mother was commenced on insulin after her previous gestation. Her last gestation resulted in therapeutic abortion as amniocentesis had shown the baby to be affected with leucodystrophy.
The mother had previously been treated with fluoxetine for depression. She smoked 10 cigarettes per day. Her diabetes was poorly controlled as was evident from glycosylated haemoglobin concentration of 10.1% in the middle trimester. An ultrasound scan was performed at 15 weeks of gestation with amniocentesis to exclude metachromatic leucodystrophy. At 26 weeks of gestation, she was receiving biphasic isophane insulin (PenMix 30/70) 24 units in the morning and 20 units at night. Her blood glucose concentrations were between 6 and 8 mmol/L. The results of the amniocentesis were normal. In view of the past poor obstetric history, an elective caesarean section at 38–39 weeks was arranged. However, she was admitted in labour at 36 weeks of gestation. Four hours later, her infant son was born spontaneously. He cried at birth. There were no cardiotocographic abnormalities in labour. The umbilical cord was not twisted around the infant's body or limbs. There were no placental infarcts. The infant had hypoglycaemia (0.6 mmol/L) which was treated with intravenous glucose.
An area of necrosis on the left forearm and hand was noted by the delivery team and a plastic surgical opinion was sought. Examination of the forearm revealed an oblique demarcation between healthy looking skin and brown discoloured forearm and hand. This was distal to the junction of the upper and middle thirds of the forearm. There was no groove or depression at the level of the demarcation and no constriction bands on any of the other limbs. The forearm and hand felt soft and jelly-like. They were of comparable size to the opposite limb and did not bleed on pinprick. The appearance of the forearm and hand suggested that the limb had been ischaemic for several weeks in utero (Fig. 1).
The baby had a cardiac murmur. Both axillary and femoral pulses were palpable. The electrocardiogram and echocardiogram were normal. A consultation by the paediatric cardiology team excluded any structural cardiac abnormality, and a distal vascular occlusion or embolus was suggested as the cause of the peripheral gangrene. The infant's haemoglobin concentration was 15.4 g/dL and platelet count was 402 × 109/L. A radiograph of the forearm showed gas in the soft tissues with contraction deformities of the fingers.
It was apparent that the involved part of the limb was not salvageable. Antibiotics were prescribed and at the age of seven days the infant underwent amputation of the forearm. It was noted that the level of the demarcation varied in different tissue planes. The amputated arm was incinerated at the parents' request and hence was not submitted for histological examination. The infant made an uneventful recovery and three months later was seen to be moving the stump satisfactorily.
Peripheral ischaemia and gangrene in a neonate is rare, with fewer than a hundred cases reported in the literature. Ischaemia of a limb is manifest at birth or within a few hours of delivery1. Various conditions have been implicated as a cause for the ischaemic episode. Thromboembolism may be a cause and may be associated with maternal diabetes2–8. In most instances, the diabetes was poorly controlled. Ours is the ninth case reported in the literature. Other associations with ischaemia of a limb include pregnancy-induced hypertension and oligohydramnios1, pressure from the maternal pelvis3, cold injury9 and intimal agglutination10.
Venous thrombosis and thromboembolism are well established complications in infants of a diabetic mother12. In their review of 4000 autopsies of newborn infants, Oppenheimer and Esterley11 found 45 infants with venous thrombosis, of whom 13 were infants of a diabetic mother. Renal and adrenal veins were most frequently involved, and none of the 13 cases had peripheral gangrene. The first published report of peripheral gangrene in an infant of a diabetic mother was by Valderrama et al. in 19725. They suggested that the adrenal vein was the probable source of the embolus to the right brachial artery through a potent foramen ovale. Carr et al.12 suggested that substances from a placental infarct were carried into the blood of fetus through a potent foramen ovale to the right ulnar artery and right internal carotid artery14. Ward7 reported an infant of a diabetic mother with cerebral arterial thrombosis and renal vein thrombosis. Hensinger3 suggested that pressure from the maternal pelvis in a prolonged labour may induce venous thrombosis of a limb. In this case, maternal diabetes was considered to have increased the susceptibility to venous thrombosis. Wiseman et al.8 noted axillary artery occlusion by arteriography. Their patient went on to develop gangrene of the forearm despite having a thrombectomy within 17 hours of delivery. According to Blank et al.13, arterial occlusion could be a result of a hypercoagulable state. Van Allen et al.6 also proposed coagulation abnormalities in infants of diabetic mothers as the likely cause for arterial as well as venous thrombosis. Decreased production of prostacyclin14 and low levels of antiplasmin15 have been shown in these infants. There have been two reported cases of peripheral ischaemia in infants of diabetic mothers where abnormal coagulation studies have been demonstrated2,6. In two other cases, coagulation studies were normal7,8. The importance of doing coagulation studies in all such cases has been emphasised6.
Treatment of these cases consists of maintaining fluid balance, preventing infection by keeping the area clean and giving antibiotics, anticoagulation and thrombectomy. In cases of established gangrene, amputation is required. It has been suggested that time should be allowed for the ischaemic area to demarcate in cases of dry gangrene before proceeding to amputation3. Similar views have been presented by Van Allen et al.6 and Turnpenny et al.1. Their argument is that non-operative treatment helps to minimise tissue loss and to retain limb function. If gangrene is not well established and ischaemia appears to be of recent onset, anticoagulation and thrombectomy should be considered. Hsi et al.4 also suggest in some cases fasciotomy to prevent massive muscle necrosis of the forearm.
The authors would like to thank the help and advice of Dr A. I. Traub (Consultant Obstetrician, Royal Maternity Hospital, Belfast) in the preparation of the manuscript.